1. Introduction
A number of recent studies have addressed challenges in today's academic job market, including in the disciplines of anthropology (Kawa et al. Reference Kawa, Clavijo Michelangeli, Clark, Ginsberg and McCarty2018, Speakman et al. Reference Speakman, Hadden, Colvin, Cramb, Jones, Jones, Lulewicz, Napora, Reinberger, Ritchison, Edwards and Thompson2018), anthropological archaeology (Speakman et al. Reference Speakman, Hadden, Colvin, Cramb, Jones, Jones, Kling, Lulewicz, Napora, Reinberger, Ritchison, Rivera-Araya, Smith and Thompson2017), English (Colander and Zhuo Reference Colander and Zhuo2014), history (Ruediger Reference Ruediger2019), economics (Chen et al. Reference Chen, Liu and Billger2012), and the biomedical sciences (Ghaffarzadegan et al. Reference Ghaffarzadegan, Hawley, Larson and Xue2015). A common finding is that the “pedigree” of an institution or department from which an individual received their PhD is an important factor structuring one's opportunities in the academic job market after graduation (Clauset et al. Reference Clauset, Arbesman and Larremore2015, Warner and Clauset Reference Warner and Clauset2015). (See also The Ph.D. Placement Project at The Chronicle of Higher Education for a number of perspectives: https://www.chronicle.com/blogs/phd.)
Our commentary contributes to this literature, and to the field of linguistics, by shedding empirical light on the nature and degree of academic hierarchy in linguistics. Based on publicly available information collected on faculty at all PhD-granting linguistics programs across the US and Canada (section 2), we use social network and heat map visualizations to demonstrate the existence of an extraordinarily strong and relatively stable hierarchy of programs whose graduates dominate the linguistics academic job market (section 3). Our shared dataset, the supplemental material to this article, will interest both prospective graduate students and established academics, and we hope will help spur many further analyses. To provide one example, we chart the gender distributions of faculty in, and faculty graduates of, the most dominant programs. Some are not well-balanced in this respect, a pattern that likely contributes to gender imbalance in the field overall (section 4).
2. Methods
We used the Linguistic Society of America (LSA)'s Directory of Linguistics Programs and Departments (LSA n.d.) to generate a comprehensive list of PhD-granting linguistics departments and programs across the US and Canada: 63 in total (52 in the US, 11 in Canada). From January through March 2019, our research team visited the web page of each department or program to collect publicly available information about each of the 788 total tenured and tenure-track (TT) faculty members employed in these programs. (We occasionally consulted additional sources such as curricula vitae published on an individual faculty member's web page, ProQuest, or departmental alumni web pages.) The categories of data we collected are listed below, with additional explanation given as needed. Our complete database is included as supplementary material to this commentary.
• Name
• Inferred Gender — Self-reported gender designations were not available, and we therefore employed a binary male/female gender classification. This was done by checking university faculty page biographies and personal websites for pronouns; if none were available, we made judgments based on the individual's name and picture. We readily admit the imperfections of this approach, including its inability to account for the fact that not all individuals subscribe to a binary gender category and that one's gender identity may change over time (e.g., since time of hire). However, we view the category of inferred gender as an improvement over the unmarked gender category that is ascribed in other, similar studies.
• Academic Rank — We limited our sample to the tenured and tenure-track (TT) ranks of Assistant, Associate, or (Full) Professor (including named and endowed professorships), which are the job titles that are consistently listed across department web pages. We excluded Emeritus Professors and all non-tenure-stream titles (e.g., Adjunct and Postdoctoral positions, Visiting Assistant Professors, and Lecturers).
• Source of PhD — This field includes PhDs obtained from US and Canadian institutions only; doctoral degrees obtained elsewhere are listed here as “Foreign” and elaborated upon in a separate “Foreign” field (see below). Two faculty members with more than one PhD are classified together into a separate grouping called “Multiple.” We did not systematically identify the source department conferring PhDs, but note that a few individuals in our database do have PhDs from fields other than linguistics (philosophy, computer science, anthropology, etc.). Thus, not all “source” institutions in the database are one of the PhD-granting schools that form the basis of our analysis. This also means that the database includes more “source” than “destination” institutions.
• Year PhD Awarded
• Destination University — This field denotes the US or Canadian university at which a scholar held a TT faculty position (as of Winter/Spring 2019).
• Destination University Location — Whether the destination university is in the US or Canada.
• Foreign Source of PhD — For PhDs granted by university sources outside the US or Canada.
• Country of PhD — Specific country for “Foreign” (i.e., non-US and non-Canada) PhD sources.
We believe that our database captures the most complete listing of academic linguists in tenure-track (TT) positions at PhD-granting institutions (PGIs) across the US and Canada as of Winter/Spring 2019. We have selected this dataset because it represents the primary producers (and, probably, consumers) of academic linguists in the US and Canada. That said, we acknowledge that it addresses only a single, albeit highly idealized, career path among many options, as linguistics PhDs are also employed in TT positions at non-PGIs and in programs other than linguistics (indeed, the first author of this article falls into both of those categories), or work outside of academia entirely.
3. A relatively stable oligarchy
Most individuals working in the field of linguistics have an intuitive sense of who the “major players” are amongst PhD-granting linguistics departments. Our analysis demonstrates that the frequently-perceived hierarchy of linguistics programs is indeed correct. Following Speakman et al. (Reference Speakman, Hadden, Colvin, Cramb, Jones, Jones, Kling, Lulewicz, Napora, Reinberger, Ritchison, Rivera-Araya, Smith and Thompson2017) and Speakman et al. (Reference Speakman, Hadden, Colvin, Cramb, Jones, Jones, Lulewicz, Napora, Reinberger, Ritchison, Edwards and Thompson2018), we frame the total number of faculty positions in an academic sphere as an economic market. Graduates compete in this market for a finite number of academic jobs, and linguistics PGIs can be ranked according to their market share, a metric we adopt from Speakman et al. (Reference Speakman, Hadden, Colvin, Cramb, Jones, Jones, Kling, Lulewicz, Napora, Reinberger, Ritchison, Rivera-Araya, Smith and Thompson2017), defined as the percentage of TT academic positions a PGI's graduates attain over a certain period of time.
One method of visualizing patterns in the linguistics job market is to use social network (chord) diagrams. Circles (nodes) in our analyses represent individual programs and their outgoing arrows demonstrate where, within the linguistics PGI market, their graduates have obtained jobs. Figure 1 shows the complex web of all “source” and “destination” programs, while Figure 2 narrows the set to those with the highest all-time market share. With no clear natural breakage line, we chose the somewhat arbitrary standard of schools with fifteen or more placements (coincidentally totalling fifteen schools), a group which comprises about a quarter of the total programs in the US and Canadian market (15/63, 23.8%).Footnote 1 In both figures, node and text label size as well as relative arrow thickness all correlate to the number of placements a department has made. Self-referential links correspond to placements of alumni who return to their own PhD-granting institution as TT faculty.
Kawa et al. (Reference Kawa, Clavijo Michelangeli, Clark, Ginsberg and McCarty2018) invite us to think about links between nodes in terms of reciprocity. Links that flow both ways between two institutions indicate a reciprocal relationship, and when these links are same-sized, the relationship is also symmetrical. Links that flow in only one direction, or differ greatly in the size of flows between two schools, however, indicate a relationship that is non-reciprocal and/or asymmetrical. In the latter case, the difference in arrow size indicates the degree of asymmetry. Figures 1 and 2 demonstrate that the patterned relationships that have built up over time between PGI programs are far from reciprocal. Instead, graduates of top market-share programs have been hired at both other top-share programs and at lower-share programs, while the lowest-share programs are primarily consumers, rather than producers, of faculty from higher-share programs.
Our second visualization, Appendix A,Footnote 2 is a heat map indicating each program's percent of the overall market share from highest (darkest) to lowest (lightest), including “Foreign” institutions, which are grouped in aggregate.Footnote 3 Here the distribution is broken down by decade to show temporal trends; the most recent span under study (2010–2019) may be of particular interest to readers who are prospective graduate students in linguistics. One particularly striking pattern is that a majority of linguistics programs have produced just a handful of PGI academics, or none at all.Footnote 4 Conversely, the all-time top ten programs (MIT through Cornell, excluding “Foreign”) represent an oligarchy of sorts, together placing half (50.3%, n = 369) of all US- and Canada-sourced faculty members, and one department alone, MIT, has placed 11% of all current US and Canadian faculty (87/733). Many top schools, including MIT, have also recruited heavily from their own PhD graduates, as shown by the reflexive arcs in Figure 2.
While the cadre of dominant departments has remained relatively stable over time, there are a few notable fluctuations. The University of Maryland and New York University, for example, show dramatic recent gains in overall market share, while the University of Chicago, University of California-San Diego, and Yale show recent declines. Whether these recent shifts represent a more sustained trend toward greater democratization of market share in linguistics programs is yet to be seen.
4. Gender imbalance, and leading the way
Most linguistics programs place few to no graduates in the PGI market, while a select number of programs place a majority of TT job holders in this market. This is perhaps a sobering portrait of the current academic job market, but one that is important for current and prospective graduate students to understand. However, our data also point to ways that the top programs might effect positive change in the discipline. As a brief example, we examine some facets of gender representation in linguistics placements, one of a number of problems of diversity and representation that the field must continue to address (see e.g., LSA 2019, Namboodiripad et al. Reference Namboodiripad, Hou and Occhino2019, Charity Hudley et al. Reference Charity Hudley, Mallinson and Bucholtz2020).
In its 2021 annual report, the LSA provides data showing that women outnumber men in both undergraduate and graduate linguistics programs (Figure 3), and that the rate at which women have received linguistics PhDs has eclipsed that of men over most of the past 25 years (Figure 4). Nonetheless, LSA data also suggest that in the average linguistics department, men have a small numerical lead over women at both the (Full) Professor and the Assistant (untenured) levels, even though the latter is where we might expect to see the greatest recent expansion of women hires (Figure 5).
Reporting rates to the LSA departmental directory from which the Figure 3 and Figure 5 numbers are gleaned, however, are inconsistent from year to year (LSA 2021). Significantly, the number of reporting departments dropped precipitously from 102 in 2018 to only 47 departments in 2019, and 43 in 2020 (LSA 2021). In contrast, our data capture a complete snapshot of PGI academic programs (n = 63), albeit at one moment in time (Winter/Spring 2019), and hence permit a more reliable investigation into gender distributions by rank, including at the scale of individual departments.
Figure 6 demonstrates that women indeed still trail at the Full and Assistant ranks at PGIs. Further, while it is not surprising that the (inferred) male-to-female faculty ratio varies greatly between departments, faculty of several of the all-time top 15 market-share programs (Figure 7) exhibit jarring male-to-female imbalances, especially Harvard (7:2), MIT (13:4), and UPenn (11:4).
We recognize that department composition can be slow to change given the infrequent opportunities to make new tenure-track faculty hires at most institutions. Hiring outcomes also reflect numerous factors including the annual demographics of the candidate pool itself, so a complementary take on gendered rates of academic job obtainment is provided by asking where the men and women who now hold TT jobs in linguistics PGIs received their degrees. Here again we focus on the all-time top 15 market-share programs. The source PhD programs of all current faculty are tallied in Figure 8, while Figure 9 shows only the most recent graduates (faculty who earned their PhD degrees in the most recent two-decade span, i.e., 2000–2019).Footnote 5
These figures underscore that it is possible to undo problems of the past. UPenn, for example, is a male-dominated department (Figure 7) that has nonetheless succeeded in stewarding female graduates on to PGI jobs in numbers that are cumulatively consistent with and now far exceed women's overall rate of PhD obtainment in the field, as seen in Figures 8 and 9 respectively. A market behemoth, MIT's linguistics program has undergone a major shift over time. Its gender record appears dismal when tracking its faculty graduates across all years (Figure 8): 57 current male PGI faculty have received their degree from MIT, compared to only 30 women, or a rate of nearly two to one. However, in the past roughly two decades (2000–2019) the program has changed course, producing nearly equal numbers of current male and female PGI faculty (Figure 9). Other leading market-share programs, by contrast, including Harvard, the University of California, Santa Cruz (UCSC), and the University of Maryland (UMD), continue to produce very few female graduates who go on to achieve faculty positions, a concerning trend given that the ripple effects of any gender imbalance found in the discipline's few dominant programs are likely to circulate widely across the field.Footnote 6
Viewed another way, the tremendous sway that the top market-share programs have within the job market as a whole affords these programs the ability – indeed, the responsibility – to take the lead in effecting positive change in the field's hiring patterns more broadly. We conclude by urging those working in the field of linguistics, but especially those privileged to hold tenure-track positions within its most influential programs, to seek the benefits of diversity in myriad forms when it comes to hiring new colleagues, and to enrolling, mentoring, and graduating Ph.D. students who, after all, represent the future of the discipline.
Supplementary Materials
To view supplementary material for this article, please visit https://doi.org/10.1017/cnj.2024.7.
Appendix A Linguistics PGIs ranked by number/percent of academic placements
* = source program that is not a linguistics PGI in the US or Canada